OMB Discloses 3,611 Federal AI Uses, Up 70% as Sparse Detail Fuels Oversight Concerns
Updated
Updated · The Guardian · Jun 15
OMB Discloses 3,611 Federal AI Uses, Up 70% as Sparse Detail Fuels Oversight Concerns
2 articles · Updated · The Guardian · Jun 15
Summary
3,611 active or planned AI use cases across the federal government were disclosed by OMB on April 14, marking a 70% jump from the Biden administration’s final inventory.
One-line or short descriptions leave key questions unanswered about how agencies will use AI in sensitive areas including prison classification, veterans’ crisis calls, grant screening and nuclear reactor safety.
The inventory also shows uneven governance: only one cited Justice Department case proposes public involvement, while many systems avoid stricter review because agencies do not classify them as “high impact.”
Some listed uses build on older practices or Biden-era initiatives, and not all are inherently problematic—translation tools alone account for 70 cases—but the report argues disclosure without meaningful consultation will deepen public distrust.
The piece points to Canada, France, Washington DC and California as models for stronger algorithmic risk assessments, public comment and human appeal rights as AI spreads through government.
As thousands of AIs enter federal service, what prevents a catastrophic, system-wide failure?
When a government AI makes a harmful mistake, who is ultimately held accountable?
How can citizens challenge an invisible algorithm that denies them benefits or flags them as a risk?
Federal AI Expansion 2025–2026: Rapid Growth, Security Initiatives, and the Transparency Crisis Undermining Public Trust
Overview
The report highlights the federal government's rapid response to the growing security risks posed by advanced AI models. With President Trump’s 2026 Executive Order, the administration set out a strategy to modernize and protect both government and private sector information systems by expanding AI-enabled cyber defenses. This move reflects a broader trend: as frontier AI models become more capable and essential for safeguarding critical infrastructure, collaboration between government and the private sector is increasing. The report also notes that these actions are part of a larger effort to balance innovation, security, and public trust as AI adoption accelerates across federal agencies.